Title :
Speeding-up image processing in reaction-diffusion cellular neural networks using CUDA-enabled GPU platforms
Author :
Stoica, George Valentin ; Dogaru, Radu ; Stoica, Elena Cristina
Author_Institution :
Dept. of Appl. Electron. & Inf. Eng., Univ. “Politeh.” of Bucharest, Bucharest, Romania
Abstract :
Due to their inherent architecture, the discrete time model of Cellular nonlinear networks (CNNs) for image processing are well suited candidates for efficient implementation using massively parallel architectures. This paper proposes an implementation model for GPU architectures and highlights the advantages over the CPU version, using nVidia´s CUDA platform.
Keywords :
cellular neural nets; image processing; parallel architectures; CNN; CUDA-enabled GPU platforms; GPU architectures; massively parallel architectures; reaction-diffusion cellular neural networks; speeding-up image processing; Computational modeling; Computer architecture; Graphics processing units; Hardware; Image processing; Instruction sets; Parallel processing; CUDA-enabled GPU; nonlinear image processing; reaction-diffusion CNN;
Conference_Titel :
Electronics, Computers and Artificial Intelligence (ECAI), 2014 6th International Conference on
Print_ISBN :
978-1-4799-5478-0
DOI :
10.1109/ECAI.2014.7090162
